you can do prestress analysis in ansys workbench as well.

Besides considering your literature method 2 or 3 will be more effective in LS-dyna case

Mr Jacob,

I know it's already 3 years now, but please can you share us your findings?

I'm modeling a similar case for prestressed beam using LS-DYNA.

I'm following the 3rd method; initial stressing.

Can you also share you're literature resources please?

Good question.. Please share me the best answer might you trust...

Regards…

Perhaps Aaron Freidenberg could share his experience modeling prestressed concrete in LS-DYNA ...?

Dear Jakob,

You'll find at the end of this message an extract of an LS-Dyna keywords file I made a couple of years ago. It's 9 years old so it may not be fully consistent with the latest release of LS-Dyna.

Anyway, I propose besides that a set of guidelines and the domain that you can cover with that modeling.

This modeling approach doesn't account for any prestressing loss like shrinkage or creep. If you want to cover it you have to define that by following some of your prefered standards. This approach requires to have calculated a priori the loss due to ageing and a part of the instantaneous ones. The way it is modeled doesn't account for the losses due to deviations. If you want to do so you may also need to define that and discretize your cable to applky the tensioning that is deriving.

Recom. 1: use elastic "plugs" at the ends of the cable if you model concrete with damage constitutive equations like MAT_159. If not there will be a concrete burst at the anchorages and propagate along the cable.

Recom 2: model stirrups in the concrete if there is a big curvature. On the contrary, you'll simulate a nice delamination.

Recom 3: start with a dynamic relaxation to balance the prestress and then carry out the transient dynamic part of your silmulation (I think it is not in my .k extract)

This modeling I propose is gross but efficient. I used it to carry out a calculation to have the first estimation of the behavior of a prestressed nuclear containment submitted to an aircraft crash with a generic and simplified model of aircraft (consistent with Riera diagram) for a paper that you can find on my RG page and that has been presented at the SMiRT'15.

I hope this will help.

Feel free to ask me any questions if anything is unclear.

Best regards

Guillaume

PS : extract of the .k

$# LS-DYNA Keyword file created by LS-PREPOST 2.4(Beta) - 21Oct2008(10:31)

$# Created on May-7-2009 (16:53:05)

*KEYWORD 300000000 NCPU=4

*TITLE

$# title

Analysis of Prestressed Containment Behavior under aircraft impact

*CONTROL_HOURGLASS

$# ihq qh

1 0.100000

*CONTROL_TERMINATION

$# endtim endcyc dtmin endeng endmas

800.00000

*DATABASE_BINARY_D3PLOT

$# dt lcdt beam npltc psetid

10.000000

$# ioopt

0

*DATABASE_FORMAT

$

$# iform ibinary

0 0

*PART

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

$ PARTS DEFINITIONS $

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

$# title

Liner

$# pid secid mid eosid hgid grav adpopt tmid

1 1 1 0 1

*SECTION_SHELL_TITLE

Shell Liner

$# secid elform shrf nip propt qr/irid icomp setyp

1 2 1.000000 3 3 0 0 1

$# t1 t2 t3 t4 nloc marea idof edgset

6.000000 6.000000 6.000000 6.000000

*MAT_PLASTIC_KINEMATIC_TITLE

Steel Liner

$# mid ro e pr sigy etan beta

1 0.007860 2.1000E+5 0.280000 350.00000 2100.0000

$# src srp fs vp

0.000 0.000 0.000 0.000

*HOURGLASS

$# hgid ihq qm ibq q1 q2 qb/vdc qw

1 1 0.100000 0 1.500000 0.060000 0.100000 0.100000

*PART

$# title

Containment

$# pid secid mid eosid hgid grav adpopt tmid

2 2 2 0 1

*SECTION_SOLID_TITLE

Solid Concrete Containment

$# secid elform aet

2 1

*MAT_CSCM_CONCRETE_TITLE

Containment concrete

$# mid ro nplot incre irate erode recov itretrc

2 0.002300 1

$# pred

0.000

$# fpc dagg units

45.000000 0.000 1

*PART

$

$

$

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

$ PARTS DEFINITIONS $

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

$

$

$# title

Vertical Prestressed Tendons

$# pid secid mid eosid hgid grav adpopt tmid

3 3 3 0 1

*SECTION_BEAM_TITLE

Cross Area Vertical Prestressed Tendons

$

$

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

$ SECTION DEFINITIONS $

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

$

$# secid elform shrf qr/irid cst scoor nsm

3 6

$# vol iner cid ca offset rrcon srcon trcon

0.000 0.000 0 7560.0000

1

*MAT_CABLE_DISCRETE_BEAM_TITLE

Steel Vertical Prestressed Tendons

$# mid ro e lcid f0

3 0.007860 1.9000E+5 0 8.8200E+6 1.000E+20

*PART

$

$

$

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

$ PARTS DEFINITIONS $

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

$

$

$# title

Horizontal Prestressed Tendons

$# pid secid mid eosid hgid grav adpopt tmid

4 4 4 0 1

*SECTION_BEAM_TITLE

Cross Area Horizontal Prestressed Tendons

$

$

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

$ SECTION DEFINITIONS $

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

$

$# secid elform shrf qr/irid cst scoor nsm

4 6

$# vol iner cid ca offset rrcon srcon trcon

0.000 0.000 0 7560.0000

1

*MAT_CABLE_DISCRETE_BEAM_TITLE

Steel Horizontal Prestressed Tendons

$# mid ro e lcid f0

4 0.007860 1.9000E+5 0 6.5600E+6 1.000E+20

*PART

$

$

$

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

$ PARTS DEFINITIONS $

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

$

$

$# title

Stirrups

$# pid secid mid eosid hgid grav adpopt tmid

5 5 5 0 1

*SECTION_BEAM_TITLE

Cross Area Stirrups

$

$

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

$ SECTION DEFINITIONS $

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

$

$# secid elform shrf qr/irid cst scoor nsm

5 3

$# a rampt stress

153.94000

*MAT_ELASTIC_TITLE

Steel Stirrups

$# mid ro e pr da db not used

5 0.007860 2.0000E+5 0.280000

*PART

$

$

$

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

$ PARTS DEFINITIONS $

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

$

$

$# title

Welded Wire Mesh

$# pid secid mid eosid hgid grav adpopt tmid

6 6 6 0 1

*SECTION_BEAM_TITLE

Cross Area Welded Mesh

$

$

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

$ SECTION DEFINITIONS $

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

$

$# secid elform shrf qr/irid cst scoor nsm

6 3

$# a rampt stress

1608.5000

*MAT_ELASTIC_TITLE

Steel Welded Mesh

$# mid ro e pr da db not used

6 0.007860 2.0000E+5 0.280000

*PART

$# title

Elastic Raft

$# pid secid mid eosid hgid grav adpopt tmid

7 7 7 0 1

*SECTION_SOLID_TITLE

Solid Concrete Raft

$# secid elform aet

7 1

*MAT_ELASTIC_TITLE

Concrete Raft

$# mid ro e pr da db not used

7 0.002300 36000.000 0.170000

*SET_PART_LIST_TITLE

Steel components

$# sid da1 da2 da3 da4

1

$# pid1 pid2 pid3 pid4 pid5 pid6 pid7 pid8

3 4 5 6

*CONSTRAINED_LAGRANGE_IN_SOLID_TITLE

$# coupid title

1Steel components bond with containment

$# slave master sstyp mstyp nquad ctype direc mcoup

1 2 0 1 0 2 1

$# start end pfac fric frcmin norm normtyp damp

0.0001.0000E+10 0.100000 0.000 0.500000

$# cq hmin hmax ileak pleak lcidpor nvent blockage

0.000 0.000 0.000 0 0.010000

*CONSTRAINED_LAGRANGE_IN_SOLID_TITLE

$# coupid title

2Vertical Prestressed Tendons / Raft

$# slave master sstyp mstyp nquad ctype direc mcoup

3 7 1 1 0 2 1

$# start end pfac fric frcmin norm normtyp damp

0.0001.0000E+10 0.100000 0.000 0.500000

$# cq hmin hmax ileak pleak lcidpor nvent blockage

0.000 0.000 0.000 0 0.010000

*DAMPING_GLOBAL

$# lcid valdmp stx sty stz srx sry srz

0 0.070000

*BOUNDARY_SPC_NODE

*ELEMENT_SOLID

*ELEMENT_SHELL

*ELEMENT_BEAM

*NODE

*END

By the way, I'm not ok with the thermal method.

It will generate an isotropic expansion or contraction in the 3 directions with the "sign". My concern is that prestressing if we consider that it generates contraction following the axis of the cable will generate expansion perpendiculart to it because of the Poisson effect.

Maybe they do something smarter than that and that there is not this non-physical effect.

There is another alternative. If the prestressing is used to counterbalance the effect of an external force/pressure applied on the structure by its inner stress provided by the stress you may just apply this external pressure/force with the opposite sign.

Some engineers do so in order to roughly estimate the prestressing within a nuclear containment model with shell elements. The prestressing has to counternalance the inner pressure in case of loss of cooling accident. They juste apply the design pressure on the outer part of the containment.

This way you can have an access to the stress distribution in singular thick zones. The consequences of out-of-plane shear forces will not be well modeled too...

Dear @Guillaume Hervé-Secourgeon,

Thanks for the valuable comments.

In the thermal method, a negative thermal load is applied only to tendon end nodes to simulate the tendon shrinkage in the rest of the tendon (negative strains). I agree that Poisson's effect may come into action, but in most cases tendons are modeled as beam elements which means only axial-longitudinal deformations will be considered by the solver (neglecting perpendicular ones), am I right?

Ok, I have been mistaken, I thought that it was some part of brick elements that were contracted by means of thermal loads. The way you describe may not induce a big mistake considering the Poisson effect I mentioned.

Using beams (potentially truss bars should be better as we shouldn’t increase some bending capabilities) and drivind the tensioning with a termal negative gradient has sense.

Thanks for the complements it helps my understanding of the way people may carry this modeling out.

And in fact no effect contafictory to the expected Poisson effect if modrled like this.